【StarRocks版本】例如：2.4.2
【集群规模】例如：3fe + 10be
fe日志：
2023-12-15 17:44:08,536 INFO (main|1) [StarRocksFE.start():109] StarRocks FE starting…
2023-12-15 17:44:08,546 INFO (main|1) [FrontendOptions.analyzePriorityCidrs():285] configured prior_cidrs value: 10.11.51.0/24
2023-12-15 17:44:08,551 INFO (main|1) [FrontendOptions.initAddrUseIp():235] Use IP init local addr, IP: /10.11.51.44
2023-12-15 17:44:08,806 INFO (main|1) [Auth.grantRoleInternal():770] grant operator to ‘root’@’%’, isReplay = true
2023-12-15 17:44:09,051 INFO (main|1) [NodeMgr.getHelperNodes():573] get helper nodes: [10.11.51.44:9010]
2023-12-15 17:44:09,076 INFO (main|1) [NodeMgr.getClusterIdAndRoleOnStartup():373] finished to get cluster id: 829986911, role: FOLLOWER and node name: 10.11.51.44_9010_1589857039953
2023-12-15 17:44:09,082 INFO (main|1) [BDBEnvironment.ensureHelperInLocal():348] skip check local environment because helper node and local node are identical.
2023-12-15 17:44:09,118 INFO (main|1) [BDBEnvironment.setupEnvironment():255] start to setup bdb environment for 1 times
2023-12-15 17:44:09,684 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [BDBEnvironment.setupEnvironment():265] add helper[10.11.51.44:9010] as ReplicationGroupAdmin
2023-12-15 17:44:09,691 WARN (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [StateChangeExecutor.notifyNewFETypeTransfer():45] notify new FE type transfer: UNKNOWN
2023-12-15 17:44:09,717 WARN (RepNode 10.11.51.44_9010_1589857039953(-1)|62) [StateChangeExecutor.notifyNewFETypeTransfer():45] notify new FE type transfer: FOLLOWER
2023-12-15 17:44:09,717 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [BDBEnvironment.setupEnvironment():285] replicated environment is all set, wait for state change…
2023-12-15 17:44:09,718 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [BDBEnvironment.setupEnvironment():293] state change done, current role FOLLOWER
2023-12-15 17:44:09,724 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [BDBEnvironment.setupEnvironment():297] end setup bdb environment after 1 times
2023-12-15 17:44:09,737 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [GlobalStateMgr.loadImage():1140] start load image from /home/shared/doris/meta/image/image.454895343. is ckpt: false
2023-12-15 17:44:09,737 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [GlobalStateMgr.loadHeader():1253] finished replay header from image
2023-12-15 17:44:09,738 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [NodeMgr.loadLeaderInfo():1067] finished replay masterInfo from image
2023-12-15 17:44:09,739 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [BDBHA.addHelperSocket():245] add 10.11.51.46:9010 to helper sockets
2023-12-15 17:44:09,739 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [BDBHA.addHelperSocket():245] add 10.11.51.45:9010 to helper sockets
2023-12-15 17:44:15,467 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [LocalMetastore.loadDb():318] finished replay databases from image
2023-12-15 17:44:18,503 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [EsRepository.registerTable():66] Thread UNKNOWN 10.11.51.44_9010_1589857039953(-1): register a new table [Table [id=2364097, name=t1, type=ELASTICSEARCH]] to sync list
2023-12-15 17:44:18,503 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [EsRepository.registerTable():66] Thread UNKNOWN 10.11.51.44_9010_1589857039953(-1): register a new table [Table [id=2243108, name=t2, type=ELASTICSEARCH]] to sync list
2023-12-15 17:44:18,503 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [EsRepository.registerTable():66] Thread UNKNOWN 10.11.51.44_9010_1589857039953(-1): register a new table [Table [id=46745878, name=t3, type=ELASTICSEARCH]] to sync list
2023-12-15 17:44:18,504 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [Load.loadLoadJob():944] finished replay loadJob from image
2023-12-15 17:44:21,394 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [GlobalStateMgr.loadAlterJob():1268] finished replay alterJob from image
2023-12-15 17:44:21,400 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [CatalogRecycleBin.loadRecycleBin():994] finished replay recycleBin from image
2023-12-15 17:44:21,415 INFO (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [VariableMgr.loadGlobalVariable():546] finished replay globalVariable from image
2023-12-15 17:44:21,423 ERROR (UNKNOWN 10.11.51.44_9010_1589857039953(-1)|1) [StarRocksFE.start():169] StarRocksFE start failed
java.lang.IllegalStateException: Cluster must be default_cluster
at com.google.common.base.Preconditions.checkState(Preconditions.java:510) ~[spark-dpp-1.0.0.jar:?]
at com.starrocks.server.LocalMetastore.loadCluster(LocalMetastore.java:3966) ~[starrocks-fe.jar:?]
at com.starrocks.server.GlobalStateMgr.loadImage(GlobalStateMgr.java:1163) ~[starrocks-fe.jar:?]
at com.starrocks.server.GlobalStateMgr.initialize(GlobalStateMgr.java:859) ~[starrocks-fe.jar:?]
at com.starrocks.StarRocksFE.start(StarRocksFE.java:115) [starrocks-fe.jar:?]
at com.starrocks.StarRocksFE.main(StarRocksFE.java:67) [starrocks-fe.jar:?]
您的详细操作步骤是什么样的？另外2.4版本已经不维护，建议升级2.5最新版本
1赞
先升级be，再升级fe，再broker。直接升级到2.5也报同样的错误。
解决了。
原因是这个集群是从doris升级过来的，原来集群里有一个废弃的cluster。在2.4.0版本之后，fe启动的时候会判断isDefaultCluster
https://github.com/StarRocks/starrocks/pull/7653/files#diff-1b0d077c3f2a45922be5fa50d2221e0e7cb4883e224636656434d449ef37ea9cR3548
由于image文件中有废弃的cluster，导致fe无法启动。
解决方案是：
回退fe到2.2，然后执行ADMIN SET FRONTEND CONFIG (“disable_cluster_feature”=“false”); drop cluster xxx;
然后重启fe触发checkpoint重新刷新image文件后，再升级到2.4可以启动通过。